Intelligent construction site management system

Abstract

The invention provides an intelligent construction site management system. The intelligent construction site management system comprises cloud deployment, edge deployment and end deployment. The cloud deployment comprises a data receiving and processing module, a service fusion and rule making module and a data storage and management module, and is responsible for data receiving, service fusion, rule making and storage and management of mass data. And the edge deployment comprises a data acquisition and primary processing module, a self-adaptive algorithm application module and an edge calculation module, and is responsible for data cleaning, screening, threshold rule dynamic adjustment and field data processing and analysis. The end deployment comprises a data acquisition and uploading module, an anomaly detection and alarm module and a feedback mechanism module, and is responsible for real-time data acquisition, anomaly detection and alarm and operation strategy adjustment. According to the system, through cooperative work of cloud, edge and end deployment, intelligentization and high efficiency of construction site management are realized, the management efficiency, the safety and the resource utilization rate are remarkably improved, and green construction and sustainable development are promoted.

Description

Technical Field

[0001] This invention relates to a construction site management system, and more particularly to an ultra-high voltage intelligent construction site management system. Background Technology

[0002] In recent years, with the rapid development of the social economy and the acceleration of urbanization, the number of infrastructure construction projects has been increasing, especially in the construction of major projects such as ultra-high voltage power projects, bridges, high-speed railways, and subways. Significant progress has been made in construction site management technology. Construction site management is a crucial link in ensuring project quality, safety, and schedule; however, traditional construction site management methods are no longer sufficient to meet the needs of modern construction. Therefore, intelligent construction site management systems have emerged as an important means to improve construction efficiency and ensure construction safety.

[0003] Traditional construction site management relies primarily on manual inspections and record-keeping, but this approach has several limitations. First, data collection and transmission are severely delayed. Construction site management involves a large amount of on-site data, such as personnel entry and exit, equipment operation, and environmental monitoring. Traditional methods, relying on manual recording and transmission, are inefficient, error-prone, and unable to achieve real-time monitoring and rapid response. Second, safety hazards cannot be detected and warned of in a timely manner. Due to the long cycles and limited coverage of manual inspections, many safety hazards cannot be detected and warned of in a timely manner, easily leading to accidents. For example, issues such as the safety of personnel working at heights, equipment failures, and the operational status of the power system are difficult to monitor comprehensively and promptly using traditional methods. Furthermore, management methods are limited, relying mainly on the experience and judgment of on-site managers, lacking scientific management methods and technical support, making it difficult to cope with complex construction environments and changing construction conditions. Finally, information silos are severe. Construction site management involves multiple departments and processes. Under traditional management methods, information flow between departments is inefficient, leading to severe information silos and hindering data sharing and collaborative management.

[0004] Furthermore, ultra-high voltage (UHV) power projects, due to their large scale, complex technology, and harsh construction environments, place higher demands on site management systems. First, high reliability is essential. The complex environment of UHV construction sites demands extremely high reliability from the management system; any system failure can have serious consequences. Therefore, a highly reliable management system capable of stable operation under various harsh conditions is required. Second, high safety is paramount. UHV projects involve high-voltage power facilities, highlighting safety issues. The management system needs strong safety monitoring and early warning capabilities to promptly detect and address various safety hazards, ensuring the safety of construction personnel and equipment. In addition, high flexibility is also necessary. Because UHV construction sites are widely distributed, the management system needs to be highly flexible, dynamically adjusting system configuration and management strategies according to different construction environments and conditions to meet real-time needs on site.

[0005] Therefore, it is necessary to propose a smart construction site management system for ultra-high voltage (UHV) projects to meet the special needs of UHV projects for site management and provide strong technical support for the smooth implementation of the projects. Summary of the Invention

[0006] To address the aforementioned problems in the existing technology, this invention proposes a smart construction site management system, comprising cloud deployment, edge deployment, and terminal deployment. The cloud deployment includes: a data receiving and processing module for receiving data from the edge deployment and terminal deployment; a business integration and rule formulation module for integrating various business scenarios, formulating perception threshold rules, and building a remote supervision application; and a data storage and management module for storing and managing massive amounts of data.

[0007] The edge deployment includes: a data acquisition and preliminary processing module, which centrally collects data from sensing devices and performs data cleaning and filtering; an adaptive algorithm application module, which dynamically adjusts sensing threshold rules and alarm strategies based on historical and real-time data; and an edge computing module, which performs data processing and analysis on-site.

[0008] The edge deployment includes: a data acquisition and upload module, which collects and uploads real-time data from the construction site through video and sensing devices; an anomaly detection and alarm module, which analyzes real-time data using preset threshold rules and adaptive algorithms, detects anomalies, and generates alarm information; and a feedback mechanism module, which provides real-time feedback and adjusts operational strategies based on the data analysis results from cloud and edge deployments.

[0009] The cloud-deployed data receiving and processing module receives data from edge and endpoint deployments and processes the data using big data analytics and machine learning algorithms. The data receiving and processing module includes a data cleaning submodule for preprocessing the raw data, including deduplication, missing value imputation, and outlier detection; a data normalization submodule for scaling all feature values ​​to the [0,1] range; a big data analytics submodule for parallel computation and analysis based on the Hadoop or Spark framework; and a machine learning submodule for modeling and predicting the data using the random forest algorithm.

[0010] The machine learning submodule uses the random forest algorithm for training and prediction, including: setting the number of trees and the maximum depth; randomly sampling from the original dataset using the Bootstrap sampling method; randomly selecting some features for optimal splitting to construct multiple decision trees; and integrating all decision trees to form a random forest model.

[0011] The business integration and rule-making module includes:

[0012] The construction progress management submodule acquires real-time progress information from the construction site through IoT devices and sensors, and generates construction progress reports.

[0013] The safety monitoring submodule monitors the safety status of the construction site in real time through cameras and sensors, automatically identifies safety hazards and generates alarm information;

[0014] The equipment management submodule monitors the operating status of equipment in real time, predicts potential equipment failures, and generates early warning information.

[0015] The threshold rule formulation submodule formulates threshold rules for various sensing devices based on historical data and expert experience.

[0016] The data storage and management module includes:

[0017] Distributed databases and file systems are used for data storage, ensuring high availability and disaster recovery capabilities.

[0018] The data backup submodule employs an off-site backup strategy and performs regular data backups.

[0019] The access control submodule ensures data access security through access control lists and role management;

[0020] The data encryption submodule uses the AES-256 algorithm to encrypt and store data, and uses the SSL / TLS protocol to encrypt and transmit data.

[0021] The data from the edge-deployed sensing devices includes temperature, humidity, pressure, vibration, and video monitoring data; and the data collected by the sensing devices is transmitted to the edge agent via a wireless or wired network.

[0022] The adaptive algorithm application module includes:

[0023] The machine learning module is used to train models using historical data and update them based on real-time data.

[0024] The data analysis module is used to analyze the collected data and generate analysis reports and alarm information;

[0025] A dynamic threshold adjustment unit for sensing devices dynamically adjusts the threshold values ​​of the sensing devices.

[0026] The alarm strategy dynamic adjustment unit is used to dynamically adjust the alarm strategy based on the model prediction results.

[0027] The new threshold of the sensing threshold dynamic adjustment unit is:

[0028] New threshold = α·historical threshold + (1-α)·real-time data average value, where α is an adjustment coefficient with a value range of 0 to 1. The historical threshold is set based on historical data and expert experience, reflecting the normal range of the construction site in long-term operation. The real-time data average value reflects the current immediate condition of the construction site environment.

[0029] The alarm probability formula for the alarm strategy is:

[0030]

[0031] Model parameters β0, β1, β2...β n Trained by a machine learning algorithm, it reflects the degree of influence of each input feature on the alarm probability. Input features x1, x2, ..., x n These come from environmental parameters and construction site data collected by sensing devices.

[0032] The data acquisition and uploading module deployed at the edge collects core element information of the construction site in real time through video devices and sensing sensors, including high-definition cameras, temperature sensors, humidity sensors, pressure sensors, and vibration sensors; the data uploading module uploads the collected data to the edge deployment for preliminary processing via wireless or wired networks;

[0033] The anomaly detection and alarm module is used to analyze real-time data according to preset threshold rules and adaptive algorithms, detect and alarm abnormal situations. The threshold rules set thresholds for temperature, humidity, pressure and vibration according to the construction site environment and construction requirements. The adaptive algorithm dynamically adjusts the thresholds based on historical data and real-time data, and generates real-time alarms through sound, light and notification alarms.

[0034] The feedback mechanism module is used to provide real-time feedback and optimize operations based on the data analysis results deployed in the cloud and at the edge. The feedback mechanism module includes a data synchronization unit that synchronizes the data and analysis results deployed in the cloud, at the edge, and at the device in real time through a high-speed network.

[0035] Beneficial effects:

[0036] The smart construction site management system proposed in this invention achieves intelligent and efficient construction site management through the collaborative work of cloud, edge, and terminal deployments. The cloud deployment integrates data reception, business fusion, rule formulation, and data management, enabling the storage and analysis of massive amounts of data. The edge deployment is responsible for data acquisition, preliminary processing, adaptive algorithm application, and on-site data analysis, improving the real-time performance and accuracy of data processing. The terminal deployment collects and uploads data in real time through video and sensor devices, detects anomalies and generates alarm information, while simultaneously providing feedback and adjusting operational strategies. The overall system significantly improves the management efficiency, safety, and resource utilization of construction sites, promoting green construction and sustainable development. Attached Figure Description

[0037] The accompanying drawings, which are provided to further illustrate the invention and form part of this application, are not intended to unduly limit the invention. In the drawings:

[0038] Figure 1 This is a schematic diagram of the structure of the UHV intelligent construction site management system of the present invention. Detailed Implementation

[0039] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. The illustrative embodiments and descriptions are only used to explain the present invention and are not intended to limit the present invention.

[0040] like Figure 1 As shown, this invention provides an ultra-high voltage (UHV) smart construction site management system, including cloud deployment, edge deployment, and terminal deployment. This system utilizes IoT, cloud computing, big data analytics, and artificial intelligence technologies to achieve intelligent management and real-time monitoring of UHV project construction sites.

[0041] Example 1: Cloud Deployment

[0042] Deployed in the cloud at the UHV company, it features various microservice applications such as smart construction sites and smart installation. Its main functions include data reception and processing, business integration and rule formulation, and data storage and management.

[0043] First, the cloud system receives data from edge and endpoint deployments and processes it using big data analytics and machine learning algorithms. This process not only generates new intelligent analysis models and predictive solutions but also significantly improves the system's predictability and intelligence. Through in-depth data analysis, the cloud system can provide more accurate predictions and smarter decision support, further optimizing the management and operational efficiency of construction sites.

[0044] Secondly, the cloud system integrates various business scenarios, establishes perception threshold rules, and builds remote supervision applications to achieve automatic perception and analysis, online monitoring and statistical analysis, and intelligent identification and analysis of core production elements. These rules and applications can monitor various indicators at the construction site in real time, automatically identify and analyze potential risks and anomalies, thereby ensuring the safety and efficiency of the construction site. Furthermore, through remote supervision applications, managers can monitor and manage the construction site anytime, anywhere, improving the flexibility and effectiveness of management.

[0045] Finally, the cloud system is also responsible for storing and managing massive amounts of data, ensuring its security and integrity. In this process, the cloud system utilizes advanced data storage and management technologies to guarantee high availability and reliability. Simultaneously, through stringent security measures and strategies, the cloud system effectively prevents data leakage and loss, ensuring data security and integrity.

[0046] In summary, cloud deployment enables efficient and intelligent management of construction sites through functions such as data reception and processing, business integration and rule formulation, and data storage and management. The cloud system not only enhances the system's predictability and intelligence but also significantly improves the safety and management efficiency of construction sites through remote supervision and intelligent analysis. Through these measures, the UHV smart construction site management system can provide comprehensive and reliable construction site management support, ensuring the smooth progress of the construction process.

[0047] In the UHV smart construction site management system, the cloud-based deployment receives data from edge and end-user deployments via a secure network connection. This data includes environmental parameters (such as temperature and humidity), equipment operating status, and worker activity information. To ensure data security during transmission, the system employs encryption technology.

[0048] Upon receiving the data, the first step is data cleaning, which preprocesses the raw data. Data cleaning includes deduplication, missing value imputation, and outlier detection. Deduplication is performed quickly using a hash algorithm. Missing value imputation uses interpolation or mean imputation for continuous data, and mode imputation for categorical data. Outlier detection uses box plots or the 3σ method to identify and remove outliers. Next, the data is normalized, scaling all feature values ​​to the [0,1] interval to eliminate dimensional differences between different features.

[0049] After cleaning and normalization, the data is then subjected to big data analysis. The system employs big data processing frameworks such as Hadoop or Spark, leveraging parallel computing and analytical capabilities to process massive amounts of data. Data is stored in HDFS (Hadoop Distributed File System) to ensure high availability and reliability. Data analysis is performed in batches using the MapReduce programming model, generating statistical reports and analysis results.

[0050] Based on the processed data, the system employs the Random Forest algorithm for machine learning. The Random Forest algorithm improves model stability and accuracy by constructing multiple decision trees. During model construction, the Random Forest uses Bootstrap sampling to randomly sample from the original training set to build the training set. At each node split, a subset of features is randomly selected for the optimal split. The number of decision trees (n_estimators) is set to 100 to enhance model stability. The maximum tree depth (max_depth) is determined through cross-validation to prevent overfitting. The node splitting criterion uses information gain, selecting features that minimize impurity for splitting.

[0051] Information gain is used to select the optimal splitting feature Gain(D,A), and its formula is:

[0052]

[0053] Where Entropy(D) represents the entropy of dataset D, A represents a feature, V represents the number of values ​​for feature A, and D... v This represents a subset of feature A that takes the value v.

[0054] In the UHV smart construction site management system, the application steps of the random forest algorithm are as follows: First, initialize the parameters by setting relevant parameters for the random forest, such as the number of trees and the maximum depth. Next, randomly select multiple sample subsets from the original dataset, and randomly select some features for optimal splitting at each node of each tree. Train a decision tree for each sample subset until a preset stopping condition is met (such as maximum depth or minimum number of leaf node samples). All decision trees are then integrated to form the final random forest model.

[0055] In practical applications, real-time collected sensor data is input into a trained random forest model for real-time prediction and analysis. For regression problems, the prediction formula for random forest is:

[0056]

[0057] in, f represents the final predicted value, N represents the number of decision trees, and f i(x) represents the prediction result of the (i)th decision tree for the input feature (x). The prediction results and analysis report are then output, including risk warnings and efficiency optimization suggestions.

[0058] Model optimization is an ongoing process. The system continuously learns from new data in real-world applications, improving its adaptability and accuracy through online learning and incremental training. Model updates are based on the latest data and business requirements, with regular retraining and optimization.

[0059] Assuming the system is used to predict equipment failures at ultra-high voltage (UHV) construction sites, the implementation is as follows: First, equipment operating status data is collected, including sensor data such as temperature, pressure, and vibration, and preprocessed, including missing value imputation, outlier detection, and feature selection. Then, a random forest model is trained using historical equipment operating data, with 100 trees and a maximum depth of 10. In the real-time prediction phase, the real-time collected equipment status data is input into the trained model for failure prediction. If a potential equipment failure is predicted, the system immediately generates an early warning message.

[0060] To ensure the long-term effectiveness and accuracy of the model, new equipment operation data is collected periodically to update the model, and online learning algorithms are used to enable the model to adapt to new data changes. In this way, the UHV smart construction site management system can utilize the random forest algorithm to achieve efficient and intelligent management of construction site equipment, ensuring the system's predictability and level of intelligence.

[0061] The cloud-based system integrates various business scenarios, including construction progress management, safety monitoring, and equipment management. It adopts a microservice architecture, with each business module developed and deployed independently, and communication and data sharing between modules achieved through API interfaces. In the construction progress management module, real-time progress information from the construction site, such as task completion status, worker attendance, and equipment usage, is acquired through IoT devices and sensors. Data collection uses RFID and GPS devices to locate and track construction personnel and equipment, and transmits real-time data to the cloud via wireless networks (such as 4G / 5G). Based on the collected data, data analysis algorithms generate construction progress reports, including completed tasks, incomplete tasks, and estimated completion times. According to the construction plan and actual progress, reminders are sent for upcoming project milestones to ensure timely completion of critical tasks, with reminders sent to relevant personnel via SMS, email, or in-app notifications.

[0062] The safety monitoring module uses cameras and sensors deployed at the construction site to monitor the safety status of the site in real time. High-definition cameras provide 24 / 7 video surveillance, and the video data is transmitted to the cloud for storage and analysis. Sensors monitor the site environment for temperature, humidity, and toxic gases. Data analysis employs computer vision and machine learning algorithms to automatically identify safety hazards (such as fires and falls) and generate alarm information. The image recognition algorithm, based on convolutional neural networks (CNN), automatically detects safety hazards. When a safety hazard is detected, the system automatically generates an alarm and sends it to relevant personnel via in-app notifications, SMS, or email.

[0063] The equipment management module monitors the equipment's operating status in real time to ensure normal operation. Sensors monitor operating parameters (such as temperature, pressure, and vibration) and transmit the data to the cloud. Based on machine learning algorithms, the operating data is analyzed to predict potential equipment failures. The failure prediction model uses historical failure data to train the aforementioned random forest model to predict the probability of equipment failure. The formula is:

[0064] Where P(failure) represents the predicted probability of equipment failure, N represents the number of decision trees, and f i (x) represents the prediction result of the i-th decision tree for the input feature x.

[0065] The rule-making section establishes threshold rules for various sensing devices based on historical data and expert experience. Temperature sensors have high and low temperature alarm thresholds set, dynamically adjusted according to site conditions and construction requirements. Humidity sensors have humidity alarm thresholds set to ensure construction materials and equipment are within a suitable humidity range. Personnel activity monitoring rules define safe zones and activity paths, providing real-time alarms for violations. Personnel activity paths are tracked using GPS and RFID devices and compared to preset safe zones. When personnel enter unsafe areas or deviate from preset paths, the system automatically generates alarm information.

[0066] The remote supervision application, based on perceived data and threshold rules, enables automatic perception and analysis, online monitoring and statistical analysis, and intelligent identification and analysis of core production elements. Automatic perception and analysis automates the analysis of perceived data, generating analysis reports and alarm information. The data analysis process includes data collection, data cleaning, data analysis, and report generation. Report generation automatically generates daily, weekly, and monthly reports based on the data analysis results, including key indicators and anomalies. Online monitoring and statistical analysis monitors and statistically analyzes real-time data from the construction site, generating dynamic reports. These dynamic reports are real-time updated visual reports including data on temperature, humidity, personnel activity, and equipment operation. Intelligent identification and analysis uses computer vision and image recognition technology to automatically identify safety hazards and anomalies, generating alarm information. The image recognition algorithm, based on a convolutional neural network (CNN) image recognition algorithm, automatically detects safety hazards. When a safety hazard is detected, the system automatically generates an alarm and sends it to relevant personnel via in-app notifications, SMS, or email.

[0067] In terms of data storage and management, the cloud system uses distributed databases (such as HBase) and file systems (such as HDFS) to ensure high availability and disaster recovery capabilities. To ensure data security and integrity, the system performs regular data backups using an off-site backup strategy. Data backups are performed daily, retaining the most recent month's backups to ensure rapid recovery in case of data loss or corruption. A rapid data recovery mechanism is also incorporated into the system design to ensure rapid recovery in the event of data loss or corruption, maintaining data continuity and availability.

[0068] In terms of data management, access security is ensured through Access Control Lists (ACLs) and role management. Different user roles (such as administrators, data analysts, and field operators) have different access permissions. Read and write permissions are controlled based on user roles and data sensitivity to ensure secure data use. To further protect data, the system employs encryption algorithms for the encrypted storage and transmission of sensitive data. Specifically, the AES-256 encryption algorithm is used to encrypt data, ensuring its security. During data transmission, SSL / TLS protocols are used to encrypt data, preventing leakage and tampering during transmission.

[0069] Through the above-described technical implementation scheme, the UHV intelligent construction site management system of the present invention can achieve efficient and intelligent management of construction sites, ensure data security and integrity, and improve the system's predictability and intelligence. The system not only possesses efficient data storage and management capabilities but also protects the confidentiality and integrity of data through multi-layered security measures. Data backup and recovery mechanisms provide additional protection for the system, ensuring data availability under various circumstances.

[0070] Example 2: Edge Deployment

[0071] The edge deployment is located on-site, centrally collecting data from sensing devices via an edge proxy, uploading edge computing results based on sensing threshold rules, and receiving cloud-based business commands to coordinate with on-site devices. Its main functions include data acquisition and preliminary processing, adaptive algorithm application, and edge computing. Through the edge proxy, the system can collect data from on-site sensing devices, perform preliminary processing and filtering to reduce data transmission latency and improve response speed.

[0072] In terms of adaptive algorithm application, the edge system introduces adaptive algorithms to dynamically adjust perception threshold rules and alarm strategies based on historical and real-time data. The adaptive algorithm includes a machine learning module and a data analysis module. Through continuous learning and optimization, it can achieve optimal detection and alarm strategies for different construction site scenarios. Specifically, the machine learning module trains the model using historical data and updates it based on real-time data, while the data analysis module analyzes the collected data and generates analysis reports and alarm information.

[0073] Edge computing performs calculations on-site, reducing reliance on cloud computing resources and thus improving overall system efficiency. By processing and analyzing data at the edge, the system can respond more quickly to on-site situations, providing real-time feedback and optimization suggestions. This approach not only improves data processing efficiency but also enhances the system's real-time performance and flexibility.

[0074] The specific implementation plan is as follows:

[0075] Edge deployment refers to system deployment at the engineering site. It involves centrally collecting data from sensing devices via an edge agent, uploading edge computing results based on sensing threshold rules, and simultaneously receiving cloud-based business commands and coordinating with on-site devices. This solution offers significant advantages in improving data processing efficiency and response speed.

[0076] For data acquisition, sensing devices include temperature sensors, humidity sensors, pressure sensors, vibration sensors, and video surveillance equipment. Temperature sensors collect real-time temperature data from the construction site, humidity sensors collect real-time humidity data, pressure sensors monitor pressure changes in equipment and structural components, vibration sensors monitor vibrations in equipment and structural components, and video surveillance equipment provides video monitoring data of the construction site. This data is transmitted to the edge agent via wireless networks (such as Wi-Fi, 4G / 5G) or wired networks to ensure real-time performance and accuracy.

[0077] In the initial processing, the collected data is first cleaned, mainly including deduplication, missing value imputation, and outlier detection. The purpose of data cleaning is to ensure the integrity and accuracy of the data, eliminating noise and invalid information. Next is data filtering, where data is filtered according to perception threshold rules to retain key data, thereby reducing data transmission latency. Perception threshold rules set thresholds for parameters such as temperature, humidity, pressure, and vibration based on the construction site environment and requirements. Data exceeding these thresholds is marked as critical data, ensuring the system focuses on data that has a significant impact on the construction site.

[0078] This edge deployment approach enables the system to rapidly process large amounts of sensor data on-site and upload critical data to the cloud as needed, ensuring efficient and timely data processing. Furthermore, edge deployment can receive cloud-based business instructions and coordinate with on-site equipment, achieving more intelligent on-site management and control. This deployment method not only improves the system's flexibility and response speed but also significantly enhances the safety monitoring capabilities and management efficiency of the construction site.

[0079] Furthermore, the application of adaptive algorithms is crucial in construction site management systems. Through the introduction of algorithm modules, the system can continuously learn and optimize, achieving the best detection and alarm strategies for different construction site scenarios. The algorithm modules mainly include machine learning and data analysis modules, which work together to improve the system's accuracy and adaptability.

[0080] The machine learning module trains the model using historical data and updates it based on real-time data. The model training employs a random forest algorithm, resulting in a model that accurately reflects various conditions at the construction site. For model updates, an online learning algorithm is used, enabling the model to continuously adjust and optimize parameters based on real-time data, thereby improving its accuracy and adaptability. This continuous model update ensures that the system maintains high-efficiency detection and alarm performance even when facing environmental changes.

[0081] Through the machine learning module, the site management system can dynamically adjust the threshold rules and alarm strategies of the sensing devices.

[0082] The purpose of dynamically adjusting the sensing threshold rules and alarm strategies is to better adapt to different construction site scenarios and environmental changes, thereby improving the accuracy of detection and the timeliness and accuracy of alarms. Regarding threshold rule adjustment, a formula is introduced...

[0083] The new threshold, calculated as α * historical threshold + (1-α) * average real-time data, achieves this goal. In this formula, α is an adjustment coefficient, ranging from 0 to 1. The historical threshold, set based on historical data and expert experience, reflects the normal range of the construction site during long-term operation, while the average real-time data reflects the immediate condition of the current site environment. The magnitude of the adjustment coefficient α determines the weight of the historical threshold and the average real-time data in the new threshold. When α is large, the new threshold primarily relies on the historical threshold, suitable for construction sites with relatively stable environments; when α is small, the new threshold responds more quickly to changes in real-time data, adapting to construction sites with dynamically changing environments.

[0084] In the specific implementation process, it is first necessary to collect environmental parameter data of the sensing device in real time, then calculate the average value of the real-time data, then calculate the new threshold according to the formula, and finally apply the new threshold to the sensing device to adjust its detection range.

[0085] Regarding alarm strategy adjustments, dynamic adjustments are made based on model prediction results to improve the accuracy and timeliness of alarms. Alarm strategy formula:

[0086]

[0087] This is achieved through a logistic regression model. The model parameters β0, β1, β2, etc., are obtained through machine learning algorithm training and reflect the degree of influence of each input feature on the alarm probability. The input features are x1, x2, ..., x... n The data comes from environmental parameters and site data collected by sensing devices, such as temperature, humidity, vibration, and pressure. The formula obtains the alarm probability by weighted summation of the input features and then transforming them using the sigmoid function.

[0088] In high-risk areas, the model parameter β can be adjusted to a higher value to improve alarm sensitivity; in low-risk areas, the model parameter β can be adjusted to a lower value to reduce false alarms. In practice, environmental parameter data of the construction site needs to be collected in real time first. Then, the real-time data is input into a trained logistic regression model to calculate the alarm probability. The alarm strategy is dynamically adjusted based on the calculated alarm probability, and an alarm probability threshold is set, for example, 0.8. When the calculated alarm probability exceeds 0.8, an alarm message is generated; otherwise, only the data is recorded without triggering an alarm.

[0089] Assuming the system is used to monitor temperature, humidity, and equipment vibration at a construction site, the specific implementation process is as follows: First, temperature, humidity, and vibration data are collected in real time, and then the average value of these data is calculated. The α value is set to 0.7, and the new temperature threshold is calculated as: New Temperature Threshold = 0.7 * Historical Temperature Threshold + 0.3 *

[0090] The real-time average temperature is calculated, and a new temperature threshold is applied to the temperature sensor to adjust its detection range.

[0091] For alarm strategy adjustments, after real-time collection of temperature, humidity, and vibration data, this data is input into a trained logistic regression model to calculate the alarm probability. An alarm probability threshold of 0.8 is set; when the calculated alarm probability exceeds 0.8, an alarm message is generated; otherwise, the data is recorded but no alarm is triggered.

[0092] In practical applications, this method of dynamically adjusting sensing thresholds and alarm strategies not only improves the system's flexibility and adaptability but also significantly enhances the safety monitoring capabilities of construction sites.

[0093] The application of edge computing in engineering sites significantly improves overall system efficiency by employing edge computing frameworks (such as EdgeX Foundry and Azure IoT Edge) for distributed computing. Performing data processing and analysis on edge devices reduces reliance on cloud computing resources and enables rapid response to various on-site situations.

[0094] For data processing, the system performs real-time computation and analysis on edge devices, providing immediate feedback on key data. This approach not only improves response speed but also ensures data timeliness and accuracy. During this process, processed data is cached to ensure data continuity and integrity, preventing data loss due to network latency or interruptions.

[0095] For local storage and transmission, critical data and analysis results are stored locally, using a distributed database (such as Cassandra) to ensure high availability and disaster recovery capabilities. The use of a distributed database ensures the reliability and stability of data stored locally. Furthermore, data is transmitted to the cloud for further processing via a secure network connection, employing data encryption technology during transmission to ensure data security. This secure transmission method protects data from theft or tampering during network transmission.

[0096] Overall, edge computing reduces reliance on cloud resources by processing and analyzing data on-site, improving system efficiency and responsiveness. Simultaneously, secure local storage and data transmission ensure data continuity, security, and high availability. These measures enable the system to maintain efficient and reliable operation even in the face of complex conditions at various construction sites.

[0097] Through the above-described technical implementation scheme, the UHV intelligent construction site management system of the present invention can achieve efficient and intelligent management of the construction site, reduce data transmission delay, and improve the system's response speed and overall efficiency.

[0098] Example 3 Terminal Deployment

[0099] Edge deployment plays a crucial role on construction sites, collecting and uploading real-time information on key site elements through various sensing devices to support complex scenarios such as smart installation. These sensing devices include video equipment and various sensors, which not only capture high-definition video and environmental parameters (such as temperature, humidity, pressure, vibration, etc.) at the construction site, but also receive and execute edge commands to ensure the safety and efficiency of site operations.

[0100] Data acquisition and uploading is one of the primary functions of edge deployment. Real-time data collected through video and sensing devices is rapidly uploaded to the edge deployment for initial processing. This process not only ensures the timeliness and accuracy of the data but also effectively reduces reliance on cloud resources, improving the overall system efficiency and responsiveness.

[0101] In terms of on-site alarms and feedback, the edge-deployed devices possess a robust real-time feedback mechanism. Through monitoring and data analysis of the on-site situation, the system can instantly identify anomalies in smart installation scenarios, such as detecting equipment malfunctions or worker violations, and immediately generate real-time alarms. This real-time feedback not only notifies on-site personnel through audible alarms and visual cues, but also allows the devices to promptly adjust and optimize their operating strategies after receiving data analysis results from cloud and edge deployments to cope with complex and ever-changing construction site environments.

[0102] Real-time feedback is one of the key advantages of edge deployment. Through data synchronization and real-time adjustment algorithms, it enables devices to quickly respond to and adapt to changes in the field. Data synchronization ensures that data and analysis results deployed in the cloud and at the edge are transmitted to the devices in real time, while real-time adjustment algorithms intelligently optimize device operation based on the latest data and analysis results, thereby improving the overall system's flexibility and efficiency.

[0103] Specifically, in terms of data acquisition and uploading, edge deployment utilizes video and sensing devices for comprehensive monitoring of the construction site. High-definition cameras operate 24 / 7, capturing high-definition video of the construction site and transmitting the video data in real time to the edge agent for initial processing via wireless networks (such as Wi-Fi, 4G / 5G) or wired networks. Sensing devices include temperature sensors, humidity sensors, pressure sensors, and vibration sensors. These sensors collect environmental parameters such as temperature, humidity, pressure, and vibration in real time and upload them to the edge deployment via the network for further analysis and processing.

[0104] Anomaly detection and alarm are crucial functions of edge deployments. The system analyzes real-time data collected by sensing devices using preset threshold rules and adaptive algorithms to detect anomalies such as high temperatures, overpressure, abnormal vibrations, and personnel violations. Threshold rules set thresholds for various parameters based on the site environment and construction requirements, while the adaptive algorithm dynamically adjusts these thresholds based on historical and real-time data to improve the accuracy of anomaly detection. Once an anomaly is detected, the system immediately generates an alarm and sends out audible, visual, and notification alerts through on-site equipment, ensuring timely attention to the anomaly.

[0105] Feedback mechanisms ensure that devices can adjust and optimize operations in real time. After receiving data analysis results from cloud and edge deployments, devices provide real-time feedback and adjustments. Data synchronization utilizes high-speed networks to synchronize data and analysis results deployed in the cloud, edge, and endpoints in real time, using MQTT, HTTP / 2, or WebSocket protocols to ensure low latency and high reliability of data transmission. During data transmission, local caching mechanisms ensure data continuity and integrity. End devices perform real-time data processing and analysis, rapidly responding to on-site situations, and utilizing edge computing frameworks (such as EdgeX Foundry and Azure IoTEdge) for distributed computing and real-time processing.

[0106] Intelligent optimization is achieved through optimization algorithms and models. Based on real-time data and analysis results, machine learning algorithms (such as random forests) are used for optimization adjustments. The optimization model is trained and adjusted based on real-time data to improve the system's response speed and accuracy. Optimization formula:

[0107] Optimization parameters = g(historical data, real-time data)

[0108] Here, g represents the optimization function, and historical and real-time data are the input parameters. The optimization strategy is dynamically adjusted based on on-site feedback to ensure the system can adapt to different construction site scenarios and environmental changes.

[0109] Assuming the system is used to monitor temperature, humidity, and equipment vibration at a construction site, the implementation is as follows: Sensing devices collect temperature, humidity, and vibration data in real time and upload them to edge deployment for preliminary processing. Simultaneously, cameras capture high-definition video of the construction site and transmit it to the edge agent in real time. The system detects anomalies based on the collected data, such as high temperature, overpressure, and abnormal vibration. Once an anomaly is detected, the system immediately generates alarm information, providing real-time alerts through sound, light, and notification alarms. After receiving data analysis results from the cloud and edge deployments, the devices adjust and optimize their operations in real time, such as adjusting temperature thresholds and increasing equipment operating frequency. Data and analysis results from the cloud, edge, and edge deployments are synchronized in real time via a high-speed network, enabling real-time data processing and analysis on the edge devices for rapid response to on-site conditions. Based on real-time data and analysis results, optimization algorithms are used for adjustments and optimizations to improve the system's response speed and accuracy.

[0110] Through the above-described technical implementation scheme, the UHV intelligent construction site management system of the present invention can achieve efficient and intelligent management of the construction site, ensure data security and integrity, and improve the system's predictability and intelligence. This terminal deployment scheme not only enhances the accuracy and efficiency of on-site management but also provides a powerful anomaly detection and feedback mechanism, providing a solid guarantee for the safety management of the construction site.

[0111] Through the above embodiments, this invention provides a highly efficient and intelligent ultra-high voltage (UHV) smart construction site management system. This system enables comprehensive monitoring and intelligent management of the construction site, improving the predictability and flexibility of site management and ensuring the safety and efficiency of the project. Through collaborative work across cloud, edge, and terminal deployments, the system can efficiently process and analyze large amounts of data, dynamically adjust management strategies, and achieve real-time optimization and intelligent management of the construction site.

[0112] The above description is only a preferred embodiment of the present invention. Therefore, all equivalent changes or modifications made to the structure, features and principles described in the claims of this patent application are included in the scope of this patent application.

Claims

1. A smart construction site management system, characterized in that, This includes cloud deployment, edge deployment, and endpoint deployment. The cloud deployment includes: The data receiving and processing module is used to receive data from edge deployments and end deployments; The business integration and rule setting module is used to integrate various business scenarios, set perception threshold rules, and build remote supervision applications. The data storage and management module is used to store and manage massive amounts of data; The edge deployment includes: The data acquisition and preliminary processing module is used to centrally collect data from sensing devices and perform data cleaning and filtering. The adaptive algorithm application module is used to dynamically adjust the perception threshold rules and alarm strategies based on historical and real-time data. Edge computing modules are used for on-site data processing and analysis; The terminal deployment includes: The data acquisition and upload module is used to collect and upload real-time data from the construction site through video and sensing devices. The anomaly detection and alarm module analyzes real-time data through preset threshold rules and adaptive algorithms to detect anomalies and generate alarm information. The feedback mechanism module is used to provide real-time feedback and adjust operational strategies based on data analysis results from cloud and edge deployments.

2. The intelligent construction site management system as described in claim 1, characterized in that, The cloud-deployed data receiving and processing module is used to receive data from edge and terminal deployments and to process the data through big data analysis and machine learning algorithms. The data receiving and processing module includes: The data cleaning submodule is used to preprocess the raw data, including deduplication, missing value imputation, and outlier detection; The data normalization submodule is used to scale all feature values ​​to the [0,1] interval; The big data analytics submodule performs parallel computing and analysis based on the Hadoop or Spark framework. The machine learning submodule uses the random forest algorithm to model and predict data.

3. The smart construction site management system as described in claim 2, characterized in that, The machine learning submodule uses the random forest algorithm for training and prediction, including: setting the number of trees and the maximum depth; randomly sampling from the original dataset using the Bootstrap sampling method; randomly selecting some features for optimal splitting to construct multiple decision trees; and integrating all decision trees to form a random forest model.

4. The intelligent construction site management system as described in claim 1, characterized in that, The business integration and rule-making module includes: The construction progress management submodule acquires real-time progress information from the construction site through IoT devices and sensors, and generates construction progress reports. The safety monitoring submodule monitors the safety status of the construction site in real time through cameras and sensors, automatically identifies safety hazards and generates alarm information; The equipment management submodule monitors the operating status of equipment in real time, predicts potential equipment failures, and generates early warning information. The threshold rule formulation submodule formulates threshold rules for various sensing devices based on historical data and expert experience.

5. The intelligent construction site management system as described in claim 1, characterized in that, The data storage and management module includes: Distributed databases and file systems are used for data storage, ensuring high availability and disaster recovery capabilities. The data backup submodule employs an off-site backup strategy and performs regular data backups. The access control submodule ensures data access security through access control lists and role management; The data encryption submodule uses the AES-256 algorithm to encrypt and store data, and uses the SSL / TLS protocol to encrypt and transmit data.

6. The intelligent construction site management system as described in claim 1, characterized in that, The data from the edge-deployed sensing devices includes temperature, humidity, pressure, vibration, and video monitoring data; and the data collected by the sensing devices is transmitted to the edge agent via a wireless or wired network.

7. The intelligent construction site management system as described in claim 1, characterized in that, The adaptive algorithm application module includes: The machine learning module is used to train models using historical data and update them based on real-time data. The data analysis module is used to analyze the collected data and generate analysis reports and alarm information; A dynamic threshold adjustment unit is used to dynamically adjust the threshold of the sensing device. The alarm strategy dynamic adjustment unit is used to dynamically adjust the alarm strategy based on the model prediction results.

8. A smart construction site management system as described in claim 7, characterized in that, The new threshold of the sensing threshold dynamic adjustment unit is: New threshold = α·historical threshold + (1-α)·real-time data average value, where α is an adjustment coefficient with a value range of 0 to 1. The historical threshold is set based on historical data and expert experience, reflecting the normal range of the construction site in long-term operation. The real-time data average value reflects the current immediate condition of the construction site environment.

9. A smart construction site management system as described in claim 7, characterized in that: The alarm probability formula for the alarm strategy is: Model parameters β0, β1, β2...β n Trained by a machine learning algorithm, it reflects the degree of influence of each input feature on the alarm probability. Input features x1, x2, ..., x n These come from environmental parameters and construction site data collected by sensing devices.

10. A smart construction site management system as described in claim 1, characterized in that, The data acquisition and uploading module deployed at the edge collects core element information of the construction site in real time through video devices and sensing sensors, including high-definition cameras, temperature sensors, humidity sensors, pressure sensors, and vibration sensors; the data uploading module uploads the collected data to the edge deployment for preliminary processing via wireless or wired networks; The anomaly detection and alarm module is used to analyze real-time data according to preset threshold rules and adaptive algorithms, detect and alarm abnormal situations. The threshold rules set thresholds for temperature, humidity, pressure and vibration according to the construction site environment and construction requirements. The adaptive algorithm dynamically adjusts the thresholds based on historical data and real-time data, and generates real-time alarms through sound, light and notification alarms. The feedback mechanism module is used to provide real-time feedback and optimize operations based on the data analysis results from cloud and edge deployments. The feedback mechanism module includes a data synchronization unit that synchronizes data and analysis results from cloud, edge, and terminal deployments in real time via a high-speed network.

Images